Re: fco and uplight

[Jan Hollan <jhollan@amper....muni.cz>]

1) Levelling of luminaires and more:

> However, here in the
> hill  country adjusting a FCO so that it doesn't cause glare is difficult. To
> me,  lights that are fully shielded, well below the source of the light would
> be  much better. One FCO blinds me almost every morning as I drive to work.
> It's on a side street about 100 feet up an incline from the  side street where
> I get on the main road. With no other street lights  for miles, that FCO
> glares down into my dark adapted  eyes and causes safety issues for me.
>
> To me, the surroundings should be considered for every light that is
> installed.
>
> Cindy Luongo Cassidy
> Outdoor Lighting  Associates
> _www.OutdoorLtg.com_ (http://www.outdoorltg.com/)

The world-best laws on outdoor lighting are meant in such a way, that the
downward direction (0 degrees) can be aimed toward a sloped terrain. It's
a vast mistake to aim it (for a fixture with a symmetric distribution)
vertically on a long slope. Not only there is glare for drivers riding up
the hill, there is also a zebra-like pattern on the road for drivers going
down the hill (maxima at 70 degrees hit the road too steeply, I have such
a bad example just on the street I'm living on). The tilt can be as large
as the slope tilt is, the only problematic point is near the top of the
hill (drivers coming from the opposite site should not see a tilted, i.e.
glaring fixture ``lying on the road''), so the tilt has to be reduced or
avoided for fixtures which reach above the hilltop.

> It's a valid point.  Traditionally, roadway lighting recommendations have
> included just such requirements and the Texas Highway Dept. (for one) used
> to build 3-dimensional models to ensure that roadway lighting in hilly areas
> wouldn't blind drivers at lower levels.  Obviously, ramps, bridges and other
> installations where there are multi-level roadways are also a problem.
> Today, designs are done mostly by computer and I don't know if the software
> typically examines potential glare situations like that -- perhaps others
> can tell us.
>
> But, in any case, roadway luminaires are typically field adjustable meaning
> that the lamp socket position can be changed to modify the light
> distribution -- or a directional shield can be added.  That's a reasonable
> request for anyone to make.
>
> Terry McGowan

In another words, even the best FCO fixture meant for spacings 1:5.5 is
not at all the best one for situations where its luminous maxima at 70+
degrees are to be avoided instead. Asymmetric distributions or
narrower-lightcone fixtures are to be implemented for all lighting
tasks differing from a single never ending straight street. Just if
this is done, using the offer of the world market of fixtures (not just
one or two products of a single manufacturer), we can call it lighting
design. I wonder in how many cases it has been implemented already...

> we will eliminate light pollution.  Until then we will have the
> never ending attacks on fco lighting in defense of the status
> quo who design by computer with the aim of placing photons
> on all the pavement between light fixtures.  The result is
> glare, light trespass, energy waste, and sky glow.
>
> The winners: the fixture and bulb manufacturers and the power companies.
> The losers:  our health, our safety, the night sky, and common sense.
> Steve P

The reason why ``designers'' don't like FCO is that a mere replacing a
convex refractor with a flat lens for the same fixture means really some
reduction of light around 72 degrees (no dramatic one). If they work with
just one manufacturer, or even with one kind of product (changing
sometimes mirrors, but sometimes just the lenses), they cannot be
satisfied. Unfortunately, there exists no catalogue of the best performing
photometric files (meaning some fixtures in some of their possibly many
optical configurations), so the designers are in a difficult position. Any
well-solved 1:5 to 1:7 task using FCO should be published in full extent
so that another designers could learn from it.

> In my experience with flat-lens cobras mounted at 20-25', the roof of the car
> is shielding my eyes as I enter what might otherwise be a glarey area.
>
>        Gail

Sun visor helps even in those cars, where the roof fails to do the job.
Actually, the very metrics of veiling luminance and threshold increment is
understating the benefit of FCO (or optimally CIE CO in the same time,
with <= 30 cd/klm at 80 degrees). There is and mostly should be a lot of
light at 60 or 70 degrees, but this can be very easily shielded by the
driver. Maybe the veiling luminance computation should be performed just
between 0 and 16 degrees from the driver's viewpoint -- every driver can
block all light coming from higher above. With such a modified standard,
non-FS lights would seldom win even for conventional designs.


2) cd/klm and definitions:

> From: patric <patric@ghostriders...>
> Subject: Re: fco and up light
> Jan Hollan wrote:
>
> > I remind that ``full cutoff'' category differs from a fully shielded one
> > just by adding a very loose limit of 100 cd/klm at 80 degrees and above.
>
> Has a Fully Shielded definition been adopted by an expert body?

No, the 0 cd/klm limit is just common to CIE cutoff and IESNA full
cutoff categories. Because of the definition being contained in no
standards, good laws don't use it, demanding explicitly ``luminous
intensity of 0 cd per 1000 lumens at 90° and higher''. Demanding FCO in
laws is no good solution either, as the 100 cd/klm limit is difficult
to verify and they may exist products with no photometry files,
being evidently fully shielded. No law should completely exclude their
use.

> (and could someone please bring me up to speed on cd/klm?)

A bare incandescent circular bulb emitting 1000 lumens has a luminous
intensity of 80 cd (1000 / 4 pi, 4 pi is the amount of steradians all
around the bulb), if we neglect the bulb base. If it would emit just 400 lm,
its luminous intensity would be just 32 cd. However, specific luminous
intensity would be still 80 cd/klm.

For a linear source, the total luminous flux is a `pi^2 steradian' times its
maximum luminous intensity, i.e., approx. ten times (if a HPS burner
emits 1000 cd perpendicularly to it, its total luminous flux is
10 sr * 1000 cd = 10 000 cd.sr = 10 klm) (I admit I remember number 10,
but I have to check it every time by integrating 2*2pi*(cos(x))^2 from 0 to
pi/2, cos(x)^2 being equivalent to (cos(2x)+1)/2.)

Fixtures block light in some directions, and add light in another ones.
E.g., there is light from both the burner (not a circular one, mostly,
but a small linear one) and its reflection in the mirror. Typical
maxima at 70 degrees from nadir are 400 cd/klm, four times the bare
bulb amount.

0 cd/klm means up to 5.0 cd/ 10 klm, so a 100 W HPS fixture is allowed
to reach a  luminous intensity of almost five candles horizontally
and upwards -- a plenty of light to become a conspicuous light point,
being even at 200 m from you still brighter than Venus.


> > This is the reason why there is MUCH less skyglow over areas with no other
> > lights than fully shielded ones.
>
> Models that show increased uplight from FC fixtures are corrupted by the
> false assumption that more luminaires are needed for a given area,
> completely ignoring the findings of current Small Target Visibility
> research.

Still: uplight and skyglow of cloudless sky have little in common, being
not at all proportional to each other. Uplight is no quantity to bother
about when the visibility of stars or Milky Way is concerned.
Astronomers have no professional reason to care about the total amount of
uplight, if they do, they are demonstrating their incompetency.

Uplight is a metric just for an overcast sky over uniformly populated
areas with unlit areas of not more than a couple of kilometres in size (or
even just hundreds of metres in size, for low-lying clouds) bringing
light into places which should remain naturally dark, like city parks. For
larger unlit areas, again: the almost horizontally emitted light is the
main issue (hitting clouds above unlit places), not the total amount of
uplight. Summing all uplight rays regardless of their direction is like
summing all coins regardles if they are 1 cent, 10 cent or 50 cent ones --
it's nice to know how many coins we have, but what we usually need is to
know the amount of money.

jenik